geekhack Projects => Making Stuff Together! => Topic started by: cest73 on Tue, 31 August 2021, 16:05:10

Title: [Procedure] Dye sublimation ghetto style (more cost effective way to dye sub)
Post by: cest73 on Tue, 31 August 2021, 16:05:10
(Wouldn't it be splendid I make a dedicated HOWTO for others to try follow this more cost effective way of Dye Subbing?)

Key points:
-outsourced the sublimation_ink_printing to a printing_shop and thus saved some $200 for printer, sublimation papers and inks.
-all precision centering is done by hand on cap to cap basis avoiding construction of an elaborate jig thus saving even more money and material.

A. DIY labels layout
The layout i had printed in that shop for the first key set:

NOTE: the actual final layout is to be send MIRRORED to the printer don't forget that

What printing shop?
Well, you know that place where people get custom printed coffee cups and shirts, like your town-local In the case your place of residence lacks local shop you still can resort to online or a similar service. Take precaution, tho that the ordered print is compatible with polyester sublimation.

The actual keyset once it received the dye:
Things to note and care about:

1. From the start i decided that i won't be transferring the whole layout in a single pass, the idea was to do one cap at a time so i can avoid making an elaborate press - some time was saved; some time was lost. Question is: am i happy with the trade?
Effort wise no: i got quite bored for affixing each single paper pad to every single key switch, but effect wise - the end result was quite worth it for once a year.
And yes, i already made it twice (although 2nd time for a 40%).

2. For the perfect alignment there is a chain of interdependent precautions to be maintained:
- The layout needs cut marks ( I used 0.25 mm wide crosses the size of 2 mm across as can be seen on the layout )
- The cuts with the scalpel need to go straight thru the cut marks, preferably splitting them in half.
- The remaining marks in the corners are to be trimmed away with scissors to avoid ever ugly leftover marks on the key caps
- The pad somewhat smaller than the key cap face is to be meticulously placed on the cap with Kapton tape on it's back and affixed to the center as exactly as possible.

Yes you want two copies printed at the shop, or, two sets on the sheet at the very least.
B. Hair straightening press converted to a heat press

Since I skimmed on the collective press i paid the full price for the iron:
I cut a hair straightening press in two and made a flat jig - i used carpentry clamps for pressing.

The heaters would reach 160 C in free air - perfectly safe for PBT and out of question for ABS.
Next to the rig are the leftover paper pads with the Kapton tape.

C. Carpenter clamps (the softer the better) as the pressure source (i used clamps of more than 5 kg pressure force according to a precision scale)

Each clamp has some 5 kg (50 N force) or more pressing power...

The force was excessive and all care has to be taken that it is applied in a way not to harm the keycap for the while it is heated - i took extra precaution not to bend the stems once i saw it can happen quite easily if neglected.
I'd say 10 N would be more adequate for this job.

D. Meticulous positioning of the cut out pieces to the key caps and affixing it with capstan tape (5 mm wide)

For OEM, Cherry profile or other similar shallow top key caps it suffices to support the paper pad with a two layer (some 2 mm in total) of silicone sourced from an arbitrary baking utensil.
However for the deeper SA style key caps the transfer might be sub optimal (take note on the numpad "5" on the big TKM-2021). For such, deeper keys, I had to use a additional trick - the lower piece of silicone had to be cut only 1/4 the size of the upper one and taped together with Kapton tape piece.
It was quite a balancing act to place the cap in such a way the lower smaller piece be right in the deepest place of the caps "belly".
The clamp had to be carefully positioned to squeeze the switch as straight as practical to avoid extensive bending of the silicone or God forbid deforming of the PBT.

E. Ran the pressing for 5 mins

The pressing time of 5 minutes provides for really rich transfer, although some moderate bleeding is noticeable on the edges. I gradually dialed the pressing time down to 3 min 30 sec on the orthDIY and got quite a sharp and clean print.

The pressing timing was all the while ballpark +/- 20 sec .
The press in action with four switches being dye sublimated for 5 minutes

-The cushion pads with the smaller piece. They stand upside down on the picture showing the Kapton tape keeping them centered. They are used other way up and the switch is placed on top with it's face side towards the cushion (and the heating element). I eventually made two more cushions and used them in pairs interchangeably.
-The discarded paper pads with Kapton tape. The tape sticks strong enough to be reused twice maybe more, but I deliberately didn't take this chance.  Notice the cut off corners on the pads - those had the alignment marks and i did them away for aesthetic reasons.
-Notice the more narrow double layered cushions of blue silicone beneath the key caps being processed. The cushions where by chance long enough to have enough room for two adjacent switches be placed next to each other. This somewhat streamlined the heating process.
-Notice in the top right corner the silicone utensil i used to source the silicone from. The narrow cushions are the side walls and the square stepped ones were cut out from the bottom section. 
-Notice the bent tip pliers near by to the right. Those where used for removing the Kapton tape from the still hot (hot! hot!) switches once off the press.

After five minutes of heating the clamp's foot would get quite warm, to avoid melting it I would turn the clamp over for each successive pass, this would also keep the back of the key somewhat cooler for the while.
For the Choc key caps i used an 12 mm washer (steel) to keep the stem out of contact to the clamp.

Despite all precautions the lower portion of the jig would still get hot after some 3 consecutive passes, so I would take a break (after some 12 ALPS key caps) and let the whole jig cool down while switched off.

For the Choc key caps i didn't take brakes as i ran only two caps at once and had the vacant clamps cool down clamping on an cold piece  of shelving.

One of my clamps got damaged by melting a foot while i was measuring the temperature of the heating elements. The material is possibly ABS like.

botched dye sub on the 5 key (not enough pressure in the middle)

botched cut lines leftovers on the extreme right side. I suspected this might happen but scraping the paper until white just was not enough - the line should have been cut off totally.

Actual layout used for the orthDIY

The result after two sessions in two days

Please kindly provide feedback and/or ask questions in the thread:
Title: Re: [Procedure] Dye sublimation ghetto style (more cost effective way to dye sub)
Post by: cest73 on Thu, 02 September 2021, 01:57:10
For those tuned to this thread early on this is a bump to be notified of the updated first post.

Good luck dyeing and have fun  :thumb:

Please do post outcomes  and any other sort of feedback here if you see fit.
Title: Re: [Procedure] Dye sublimation ghetto style (more cost effective way to dye sub)
Post by: OddField on Fri, 10 September 2021, 23:36:20
The heaters would reach 160 C in free air - perfectly safe for PBT and out of question for ABS.
Next to the rig are the leftover paper pads with the Kapton tape.

Thank you for this incredible and detailed post. I have tried to find information on dye-sub for a long time, and this is by far the most detailed documentation.

A question: POM melts at 175 C, but probably start to deform before that. Do you think it would be viable to dye sub a POM keycap?
Title: Re: [Procedure] Dye sublimation ghetto style (more cost effective way to dye sub)
Post by: cest73 on Sun, 12 September 2021, 06:49:57

Please provide feedback on whatever you do, be it POM or PBT or even ABS?

I found the data for ABS and PBT on some company's temperature chart, i guess you should try the same before you run destructive trials?

I am considering tackling ABS some day, with shorter time and less pressure, so stay tuned!